Change Details

In probedone(), for the PROBE_REPORT_LUNS case, all paths that fall to the bottom of the case `free()` the `lp` variable and set `lp` to NULL, so the test for a non-NULL value of `lp` and call to `free()` if true is dead code as the test can never be true. Fix by eliminating the whole if statement. I will mention that I considered just eliminating the `if` test and leaving the call to `free()` as that is harmless on a NULL pointer. That would somewhat "future-proof" the code in case some path is later added that _doesn't_ `free()` the pointer. I happened upon this CID after looking at some code coverage analysis that pointed out that the call to `free()` was never being executed. Calling `free()` unconditionally would get that coverage. Two considerations prevented me from doing that: 1) the rest of the code in file seems to always test for non-NULL before calling `free()` and 2) there would be a bit of needless overhead from the call to `free()`. Any thoughts on generally preferred practice? Adding as subscribers people who have been in the area in the current version of the code. I'd appreciate review from some or all of you.

In probedone(), for the PROBE_REPORT_LUNS case, all paths that fall to the bottom of the case `free()` the `lp` variable and set `lp` to NULL, so the test for a non-NULL value of `lp` and call to `free()` if true is dead code as the test can never be true. Fix by eliminating the whole if statement. I will mention that I considered just eliminating the `if` test and leaving the call to `free()` as that is harmless on a NULL pointer. That would somewhat "future-proof" the code in case some path is later added that //doesn't// `free()` the pointer. I happened upon this CID after looking at some code coverage analysis that pointed out that the call to `free()` was never being executed. Calling `free()` unconditionally would get that coverage. Two considerations prevented me from doing that: 1) the rest of the code in file seems to always test for non-NULL before calling `free()` and 2) there would be a bit of needless overhead from the call to `free()`. Any thoughts on generally preferred practice? Adding as subscribers people who have been in the area in the current version of the code. I'd appreciate review from some or all of you.

In probedone(), for the PROBE_REPORT_LUNS case, all paths that fall to the bottom of the case `free()` the `lp` variable and set `lp` to NULL, so the test for a non-NULL value of `lp` and call to `free()` if true is dead code as the test can never be true. Fix by eliminating the whole if statement. I will mention that I considered just eliminating the `if` test and leaving the call to `free()` as that is harmless on a NULL pointer. That would somewhat "future-proof" the code in case some path is later added that _//doesn't_// `free()` the pointer. I happened upon this CID after looking at some code coverage analysis that pointed out that the call to `free()` was never being executed. Calling `free()` unconditionally would get that coverage. Two considerations prevented me from doing that: 1) the rest of the code in file seems to always test for non-NULL before calling `free()` and 2) there would be a bit of needless overhead from the call to `free()`. Any thoughts on generally preferred practice? Adding as subscribers people who have been in the area in the current version of the code. I'd appreciate review from some or all of you.